No, it won't, because the problem is the architecture of the AI, not the lack of compute. Someone needs to first figure out the problem and then write about it, then the AI can figure it out. Perhaps a greater "reasoning" architecture will be able to govern the output but what we currently have is at best a facsimile of intelligence. The real conspiracy is tptb using the populace's ignorance to make them think this (admittedly, very useful) tool is in any way a real intelligence